Endocrinology Vol. 144, No. 4 1426-1434
Copyright © 2003 by The Endocrine Society
Prenatal Programming of Reproductive Neuroendocrine Function: Fetal Androgen Exposure Produces Progressive Disruption of Reproductive Cycles in Sheep
Rachel A. Birch1,
Vasantha Padmanabhan,
Douglas L. Foster,
William P. Unsworth and
Jane E. Robinson
Laboratory of Neuroendocrinology, Department of Neurobiology, The Babraham Institute (R.A.B., W.P.U., J.E.R.), Cambridge CB2 4AT, United Kingdom; and Departments of Pediatrics (V.P.), Obstetrics and Gynecology, and Ecology and Evolutionary Biology (D.L.F.) and the Reproductive Sciences Program (V.P., D.L.F.), University of Michigan, Ann Arbor, Michigan 48109
Address all correspondence and requests for reprints to: Dr. Jane E. Robinson, Department of Preclinical Veterinary Studies, University of Glasgow Veterinary School, Bearsden Road, Glasgow G61 1QH, United Kingdom.
 |
Abstract
|
|---|
In the agonadal, androgenized ewe testosterone before birth produces a precocious pubertal rise in circulating LH and abolishes the LH surge mechanism. The present study tested two predictions from this model in the ovary-intact female: 1) prenatal androgen exposure produces early ovarian stimulation; and 2) despite early ovarian stimulation, progestogenic cycles would not occur because of the abolition or disruption of the LH surge. Pregnant ewes were injected with testosterone propionate twice per week from either d 3090 (T60 group; 100 mg/injection) or d 6090 (T30 group; 80 mg/injection) of gestation (term, 147 d). Control ewes received no injections. At birth, the androgenized and control lambs were divided into two groups: ovary-intact to determine the effects of prenatal androgen on the timing of puberty and subsequent ovarian function, and ovariectomized to assess the timing of the pubertal decrease in sensitivity to estrogen negative feedback and the subsequent increase in LH. Neonatally orchidectomized, estrogen-treated males were included for comparison of the timing of this pubertal rise in LH secretion. Neuroendocrine puberty (determined on the basis of LH increase) was advanced in the androgenized females to a similar age as in males. Repeated progesterone cycles of the same duration and number occurred in the ovary-intact ewes, and they began at the same time as for control females, thus negating both predictions. Differences appeared during the second breeding season, when reproductive cycles were either absent (T60) or disrupted (T30 group). Our findings reveal that exposure to androgens in utero causes a progressive loss of cyclic function in adulthood.
 |
Introduction
|
|---|
THE REPRODUCTIVE neuroendocrine axis of many mammals is sexually differentiated by exposure of the fetus/neonate to male steroid hormones (1, 2). Central to this is the sex-specific regulation of the network of neurons that regulate GnRH secretion from the hypothalamus. In the sheep, one of the earliest physiological manifestations of this differential regulation becomes apparent during the pubertal process, when a decrease in sensitivity to inhibitory estrogen feedback occurs. This times the first major increase in GnRH secretion, and the resultant increase in gonadotropin secretion activates the gonads. In the well studied gonadectomized lamb model, in which peripheral steroid concentrations are clamped at physiological levels by means of a constant release device (estradiol capsule sc), LH secretion remains inhibited for a period after birth. Then, in response to growth-related and environmental cues, the mechanisms controlling tonic GnRH secretion become insensitive to estrogen negative feedback, allowing LH pulses to increase in frequency. This process, termed neuroendocrine puberty occurs in young, ovary-intact females coincident with the time of onset of progesterone cycles (3). In contrast, the time of neuroendocrine puberty occurs much earlier in the male, where testicular growth and development are more gradual and begin earlier than ovarian function. This differential timing of the initiation of puberty in the sheep arises from the organizational action of androgens secreted from the fetal testes during the critical period for sexual differentiation, which extends from approximately d 3090 of gestation (term, 147 d; see Refs. 4, 5, 6, 7 for reviews). Treatment of female lambs with testosterone during this critical period results in a marked advancement of the time of neuroendocrine puberty to an age that is virtually the same as that for the normal male.
Because neuroendocrine puberty is advanced by several weeks in the female lamb by prenatal androgenization in the feedback model (OVX+E), one would predict that in the prenatally androgenized, ovary-intact female, puberty would be precocious. On the other hand, this prediction may not hold, because prolonged exposure of the female fetus to testosterone during the critical period also renders the preovulatory GnRH surge system inoperative in the OVX+E model (8). If this were the case for the ovary-intact model as well, then no ovulations could occur, and repeated progesterone cycles would not be possible. Our recent studies in the ovary-intact Suffolk breed suggested that neither of these predictions is true (9). In these studies, in stark contrast to the classical findings in the ovariectomized model, the initiation of progestogenic cycles in prenatal testosterone-treated ovary-intact Suffolks [exposed to testosterone from d 3090 (T60) or d 6090 (T30) of gestation] not only occurred, but did so at the normal time. However, in this study the neuroendocrine model was not included. The present investigation was conducted to reexamine these paradoxical findings of the effects of prenatal androgen on postnatal reproductive function; namely, that the onset of neuroendocrine puberty and ovarian cyclicity temporally differ in the gonadectomized and ovary-intact models. Our approach was to make a direct comparison between the two models receiving the same prenatal treatment at the same time and differing only in postpubertal ovarian/steroid treatment (ovariectomized and estradiol treated or ovary intact). We also extended our knowledge of reproductive function in the ovary-intact ewe by determining whether these animals could produce normal progestogenic cycles during their first and second breeding seasons.
 |
Materials and Methods
|
|---|
Animals and steroid treatments
Studies were performed on 7 male and 52 female sheep of the Poll Dorset breed born between February 20 and March 13, 1998. Thirty-seven of the females were exposed to exogenous androgens during fetal life by injecting the mothers twice weekly with testosterone propionate (TP; in 1 ml vegetable oil, im; Sigma-Aldrich, Poole, UK) as described in a previous study (10) for either 60 d (n = 14; d 3090 of pregnancy; T60 group) or 30 d (n = 23; d 6090 of pregnancy; T30 group). A lower concentration and shorter duration of TP were used in the T30 lambs (80 mg/injection) than in T60 lambs (100 mg/injection) in an attempt to produce less masculinized females. The remaining 15 female lambs served as controls and received no in utero treatment. At birth, 7 untreated males were immediately castrated by scrotal ligation with a rubber band. The untreated and androgenized females were randomly assigned to 1 of 2 groups, either gonadectomized or gonad intact. At 5 wk of age, 7 T60, 16 T30, and 7 control female lambs were ovariectomized via midventral laparotomy. Immediately after gonadectomy, all male and female lambs were implanted sc with a 3-cm SILASTIC brand capsule (inside diameter, 3.35 mm; outside diamter, 4.65 mm: BDH, Lutterworth, UK) containing a 3-cm column of packed crystalline 17ß-estradiol (Sigma-Aldrich, St. Louis, MO) to standardize the hormonal milieu of the animals (11). The second group of T60 (n = 7), T30 (n = 7), and control female lambs (n = 8) remained ovary intact. The animals were maintained outdoors during the summer months and under shelter during the winter months in natural lighting conditions experienced in Cambridge (52°12' N). The intact male siblings were removed at weaning (6 wk of age), and the animals in these studies were never in contact with intact adult rams. Procedures were carried out under Home Office Project License PPL 80/1037.
Experimental design
Time of neuroendocrine puberty and estradiol positive feedback response in gonadectomized, estrogen-treated male, female, and androgenized female Poll Dorset sheep.
The majority of studies of sexual differentiation of reproductive neuroendocrine function have been performed in Suffolk or Suffolk cross-bred sheep in which the timing of puberty in the female is timed by photoperiod (11, 12). This experiment was performed to confirm that a similar sex difference in the timing of neuroendocrine puberty was present in Poll Dorset sheep, which is a much less seasonal breed. Sex differences in the control of the gonadotropin surge mechanism were also examined.
Neuroendocrine puberty onset.
The time of onset of neuroendocrine puberty (escape from estradiol negative feedback) was determined in gonadectomized control females, T60 and T30 females, and estrogen-treated males by measuring the concentrations of LH in samples of jugular blood collected twice per week. The time of onset of neuroendocrine puberty was defined as the first of six consecutive samples in which LH concentrations were sustained above 1 ng/ml following previously published criteria (13). Sampling began at 9 wk of age and continued until after the time of onset of neuroendocrine puberty in each group.
Estradiol positive feedback in gonadectomized control and prenatal T-treated sheep.
The ability to generate an LH surge in response to elevated estrogen was determined in gonadectomized, chronically estrogen-treated sheep during the anestrous season (March) after they had reached neuroendocrine puberty and were approximately 1 yr old. One month before the study was performed, the 3-cm estrogen implant was removed and replaced with a 1-cm estrogen implant, producing low follicular phase levels of this steroid in the peripheral circulation (12 pg/ml; Ref. 14). The presence of estrogen positive feedback was tested by inserting an additional four 3-cm estrogen implants sc to produce high follicular phase levels of 1012 pg/ml, a method that has been frequently used to elicit LH surges in normal ewes (15). After estrogen implantation, jugular blood samples were taken every hour for 35 h, beginning 10 h after the estrogen administration, to determine circulating concentrations of LH during this time. An LH surge was defined as LH values exceeding twice the average preestradiol baseline for a minimum of 6 h as described by Masek et al. (16).
Time of onset and maintenance of reproductive cycles in the ovary-intact control and androgenized Poll Dorset ewes.
The initiation and maintenance of cycles in ovarian function were determined by measuring the concentrations of progesterone in samples of jugular blood collected twice per week beginning at wk 9 of postnatal life and continuing through the second breeding season in control ewes and androgenized ewes when the animals were approximately 26 months old. The onset of ovarian cyclicity was defined as the age when progesterone concentrations first increased above 1 ng/ml for at least 3, but not more than 4, consecutive samples (spanning at least 10, but not more than 14 d and representing the normal duration of the luteal phase of the cycle) and decreased to less than 1 ng/ml between each cycle for at least 1 sample (representing the length of the normal follicular phase).
Determination of LH and progesterone concentrations.
The RIA procedure to quantify LH is based on the double antibody method initially described by Niswender et al. (17) and subsequently modified at Babraham (18). The primary antiserum was NIDDK rabbit antisheep LH and the standard was NIH-S11. Three separate assays were performed and the interand intraassay coefficients of variation were 11.4% and 10.2% respectively. The mean detection limit of the assays (2 SD from the buffer controls) was 0.3 ng/ml. Concentrations of circulating progesterone in the ovary-intact lambs were determined in jugular plasma using a Coat-a-Count RIA kit (Diagnostic Products Corp., Los Angeles, CA) which has previously been validated for use in the sheep (19). The mean detection limit of the assays was 0.2 ng/ml, and the inter- and intraassay coefficients of variation were 8.4% and 3.9%, respectively (n = 4 assays).
Statistical analysis
All statistical analyses were carried out with the aid of the Instat for MacIntosh statistical computer package. In all analyses, one-way ANOVA with Tukeys post hoc test was used. Undetectable LH and progesterone concentrations were assigned the limit of detection of the assay. All results are presented as the mean ± SEM. Significance was defined as P < 0.05.
 |
Results
|
|---|
Time of neuroendocrine puberty and estradiol positive feedback response in gonadectomized, estrogen-implanted male, female, and androgenized female Poll Dorset sheep
Neuroendocrine puberty onset.
The pattern of LH secretion in gonadectomized, estrogen-implanted control females, males and androgenized females is illustrated in Fig. 1
. The time of onset of neuroendocrine puberty (LH escape from estrogen negative feedback; indicated by an arrow) was sexually differentiated in the Poll Dorset lambs. In males, the sustained rise in LH concentrations had already occurred by the age of the first sample (9 wk), which was significantly earlier than that in female lambs (33.3 ± 0.6 wk of age; P < 0.001). Compared with these control females, the pubertal LH rise in females treated with testosterone occurred at younger ages in the T30 (21.7 ± 1.9 wk; P < 0.01) and T60 (12.4 ± 0.9 wk; P < 0.001) females.

View larger version (21K):
[in this window]
[in a new window]
|
Figure 1. Time of onset of neuroendocrine puberty in representative ovariectomized, estradiol-treated control female (top), control male (bottom), and androgenized female (middle) lambs. Blood samples were collected twice weekly from 9 wk of age. The dotted line represents 1 ng/ml. Arrows indicate neuroendocrine puberty (LH escape from estradiol negative feedback) onset in each representative lamb. , Time of onset of neuroendocrine puberty for each lamb in the group.
|
|
Surge mode of LH secretion.
Figure 2
illustrates the pattern of LH secretion after exposure to exogenous follicular phase concentrations of estrogen in gonadectomized, estrogen-implanted males, females, and androgenized female lambs. Seven of eight control females produced an unambiguous surge in LH secretion that began 12.5 ± 0.3 h (range, 12.014.0 h) after the artificial increase in estradiol was produced, and circulating LH concentrations reached a peak amplitude of 44.3 ± 3.5 ng/ml (range, 27.150.0 ng/ml). The one nonresponder was not included in the analysis because she lost her implants during the course of the study. None of the males was able to produce an LH surge in response to the estrogen signal. Fifteen of the 16 T30 lambs produced an LH surge. However, the mean time of surge onset was significantly delayed (17.1 ± 1.2 h; range, 12.026.0; P < 0.01) compared with that in control females, and the mean peak amplitude of the surges was significantly smaller (18.1 ± 3.0 ng/ml; range, 4.346.3; P < 0.001). Unlike T30 lambs, but similar to males, none of the T60 female lambs produced an LH surge.

View larger version (15K):
[in this window]
[in a new window]
|
Figure 2. LH secretion in response to a surge-inducing dose of estradiol in gonadectomized, estrogen-treated control female (left), male (right), and androgenized (middle) female ewes. Blood samples were collected hourly for 35 h beginning 10 h after estrogen administration. Values are the mean ± SEM. , The one nonresponder in the T30 group that failed to generate an LH surge in response to estrogen treatment.
|
|
Influence of prenatal testosterone on the onset and maintenance of reproductive cycles in Poll Dorset ewes
First breeding season.
The patterns of progesterone secretion in individual control, T30, and two T60 ewes are shown in Fig. 3
, and the percentage of animals exhibiting reproductive cycles during the first breeding season is presented in Fig. 4
. All control lambs initiated cycles of ovarian function during the first breeding season, and the first cycle (ovarian puberty) began, on the average, at 31.8 ± 0.7 wk of age (October 11 ± 4 d). One T30 lamb and two T60 lambs failed to generate any patterns of progesterone secretion that fit our definition of a cycle, and these were excluded from any statistical analysis. Of the remaining animals, reproductive cycles began, on the average, at 31.8 ± 0.6 wk of age in T30 lambs (October 15 ± 4 d; n = 6) and at 30.6 ± 1.3 wk of age (October 11 ± 9 d; n = 5) in T60 lambs. There was no difference in the time of ovarian puberty onset among these three groups (Fig. 5A
).

View larger version (27K):
[in this window]
[in a new window]
|
Figure 3. Plasma progesterone profiles from representative control, T30 androgenized, and T60 androgenized ewes. Blood samples were taken twice weekly beginning just before the onset of the first breeding season until the end of the second breeding season.
|
|

View larger version (34K):
[in this window]
[in a new window]
|
Figure 5. Characteristics of estrous cycles during the first breeding season in ovary-intact control and androgenized ewes.
|
|
All of the control females had regular progesterone cycles during the first breeding season, which ended on January 12 ± 12 d (Fig. 3
, individuals; Fig. 4
, all). Although one of the T30 ewes and two of the T60 ewes failed to show any cycles in progesterone secretion, cycle characteristics in the remaining androgenized animals were identical to those in the control lambs. The end of the first breeding season was similar in all three groups of lambs, with cycles stopping on January 22 ± 6 d and January 25 ± 8 d in T30 and T60 lambs, respectively. Furthermore, there was no significant difference among the three groups in the length of the first breeding season, the number of reproductive cycles, or the average peak progesterone concentration attained during each cycle in cyclic animals (Fig. 5
, BD).
Second breeding season.
Regular repeated cycles resumed in control ewes beginning, on the average, on July 27 ± 10 d and continuing until March 27 ± 8 d (Fig. 3
, individual; Fig. 4
, all). Despite five of the seven T60 ewes producing normal reproductive cycles during the first breeding season, none of these females produced any cycles during the second breeding season. In the T30 group, five of seven ewes produced progesterone cycles during the second breeding season, but in one of these females, the cycles were irregular (Fig. 3
, individual; Fig. 4
, all). Furthermore, although the time of onset of the second breeding season (August 12, 1999, ± 12 d) and the peak progesterone concentrations attained in each cycle (control, 6.1 ± 0.2 ng/ml; T30, 5.4 ± 0.9 ng/ml; Fig. 6
) in the five T30 animals were similar to those in controls, the end of the breeding season was significantly earlier (February 20 ± 19 d; P < 0.05), resulting in a shorter breeding season (control, 34.0 ± 1.1 wk; T30, 26.7 ± 2.6 d; P < 0.05; Fig. 6
) and fewer, less regular reproductive cycles (control, 13.6 ± 0.4 cycles; T30, 9.0 ± 1.8 cycles; P < 0.01; Fig. 6
).

View larger version (18K):
[in this window]
[in a new window]
|
Figure 6. Characteristics of estrous cycles during the second breeding season in ovary-intact control and androgenized ewes. *, P < 0.05; **, P < 0.01.
|
|
Comparison of the time of onset of neuroendocrine and ovarian puberty
The temporal relationship between the onset of neuroendocrine puberty in gonadectomized lambs and that of ovarian puberty in gonad-intact lambs is compared in Fig. 7
and Table 1
. As expected from other studies in our laboratories (Refs. 3 and 9 ; and Evans, N. P., and J. E. Robinson, unpublished observations), the timings of neuroendocrine and ovarian puberty were very similar in the control females. Specifically, the LH rise in the gonadectomized control animals (neuroendocrine puberty) occurred, on the average, when the animals were 33.3 ± 0.6 wk of age, and the start of progesterone cycles, indicating ovarian puberty, occurred when the ovary-intact animals were 31.8 ± 0.7 wk old. In marked contrast to the situation in the control lambs, in the animals exposed to testosterone in utero there was a significant dissociation between the increase in LH concentrations at neuroendocrine puberty in the ovariectomized animals and the onset of progestogenic cycles in ovary-intact animals (Fig. 7
and Table 1
). The delay between the start of neuroendocrine (gonadectomized) and ovarian (ovary-intact) puberty in the T30 females was 104 d (middle panel; P < 0.01), and that in the T60 females was 146 d (bottom panel; P < 0.001).

View larger version (29K):
[in this window]
[in a new window]
|
Figure 7. Cumulative data comparing the mean time of onset of neuroendocrine puberty in OVX+E ewes (line graph) and the percentage of ovary-intact ewes cycling (gray shading) during the first breeding season. The dotted line represents 1 ng/ml. The hatched circle represents the mean ± SEM age of neuroendocrine puberty (LH escape from estradiol negative feedback) in OVX+E females.
|
|
View this table:
[in this window]
[in a new window]
|
Table 1. Mean ± SEM age at the time of escape of LH from estradiol negative feedback (neuroendocrine puberty) in ovariectomized, estrogen-treated females and the time of onset of progestogenic cycles (ovarian puberty) in ovary-intact female sheep
|
|
 |
Discussion
|
|---|
The results of the present study clearly document that exposure of the female sheep fetus to androgens during a critical period for sexual differentiation of the brain progressively disrupts reproductive cycles in the adult ewe. Similar effects of androgenization on reproductive function have been described in other species. Specifically, the effects of prenatal androgen exposure have been shown to alter the timing of puberty in female rats (20), guinea pigs (21), and primates (22). In addition, pre- or neonatal testosterone treatment leads to the production of abnormal reproductive cycles in rodents (23) and primates (24), supporting the conjecture that a relatively short exposure to male gonadal hormones early in development can program the reproductive axis of the female to malfunction in later life, leading to sub- or infertility. Our findings that reproductive cycles occurred in a proportion of lambs during the first breeding season after prenatal exposure to androgens were unexpected and were not predicted from a well studied experimental model of the neural mechanisms underlying the sexually differentiated secretion of gonadotropins, the gonadectomized, estrogen-implanted sheep. In this model, animals are gonadectomized in the early postnatal period, and systemic concentrations of estrogen are clamped at physiological levels by sc implantation of a SILASTIC brand capsule containing crystalline estradiol (4, 5, 6, 7). The developing sheep of either sex is extremely sensitive to the negative feedback actions of estrogen, which profoundly suppress the circulating concentrations of gonadotropins. At the time of neuroendocrine puberty, the sensitivity of the GnRH neural network to estrogen negative feedback markedly decreases, and LH pulse frequency progressively increases to stimulate the activity of the immature gonads. Studies carried out predominantly in the Suffolk breed have shown that the time of neuroendocrine puberty is sexually differentiated (4, 5, 6), and we have confirmed this in the breed of sheep used in the current studies, the Poll Dorset. Specifically, the escape from estrogen negative feedback occurred in October in the ewe lambs when they were just over 30 wk of age, which is more than 20 wk later than in the ram lambs. The timing of this change in sensitivity is clearly programmed by in utero androgen exposure, as it occurred at a similar time in males and T60 females. The situation in animals that were exposed to slightly lower concentrations of testosterone (80 mg as opposed to 100 mg in the T60 group) for the latter half of the critical period (T30) was intermediate. This is an important finding, because day length is an important cue timing neuroendocrine puberty in ewe lambs, but not in young male or androgenized ewe lambs, of the Suffolk breed (25, 26). However, it was uncertain whether such a sexual differentiation would occur in the Poll Dorset breed, in which the ewe is much less sensitive to photoperiod influences and is, therefore, much less seasonal (Evans, N. P., and J. E. Robinson, unpublished results). Although we thought it possible that the timing of neuroendocrine puberty in the Dorset would be similar in males and females, this was clearly not the case. Thus, despite the different temporal patterns of seasonal reproduction in adults of the Suffolk and Dorset breeds, the timing of neuroendocrine puberty in both is sexually differentiated in response to in utero exposure to testosterone.
In further concordance with previous findings in the gonadectomized, estrogen-implanted Suffolk ewe model (4, 6), neither the male nor the T60 androgenized Dorset sheep displayed an LH surge in response to follicular phase concentrations of estrogen. In addition, the partially androgenized ewes (T30) of the two breeds demonstrated a similar positive feedback response, in that the LH surge was present, but delayed. Similar positive feedback responses were obtained with ovary-intact T30 ewes studied during the anestrous season (9). These observations reinforce the robust nature of these prenatal actions of androgen, which are not restricted by the characteristics of the breed or the geographic location of the sheep.
Our observations of the effects of prenatal androgen exposure on the response of the reproductive neuroendocrine system to estrogen in the gonadectomized, steroid-implanted animals led to the two, mutually exclusive predictions tested in this study. We had hypothesized that an early reduction in steroid negative feedback at neuroendocrine puberty would result in the precocious maturation of the gonads of ovary-intact androgenized animals. However, because the LH surge-generating system is absent in the gonadectomized, androgenized female, ovulation would not occur, and normal reproductive cycles would not begin. In recent studies, characterization of ovary-intact control and prenatal testosterone-treated Suffolk sheep suggested that neither of these predictions may be true (9). The results of the present study, conducted in Dorsets, extended these initial findings to another breed and, in addition, convincingly documented that the prenatal testosterone-treated animals achieve puberty at the same time as controls, a finding not in agreement with either of the predictions made from the gonadectomized model. Specifically, the majority of androgenized females initiated cyclic reproductive function. Moreover, the cycles were not precocious, and they began at the same time as those in control lambs.
How the androgenized, ovary-intact lamb can produce repeated ovulatory cycles without a positive feedback mechanism, as predicted from the neuroendocrine model (OVX+E), is unclear. The intact female is probably capable of generating an ovulatory LH signal, because not only are progestogenic cycles clearly evident in this study in the Dorset and in our earlier study in the Suffolk breed (9), but corpora lutea have been clearly identified in the ovaries of androgenized ewes (Ref. 27 ; and Birch, R., and J. E. Robinson, unpublished observations). The differences in results between the gonadectomized and ovary-intact, prenatal testosterone-treated females cannot be explained by differences in fetal programming, as both types of females were treated identically during the prenatal period. There are, however, two fundamental differences between the two groups that might begin to explain the dissociation in the timing of onset of neuroendocrine and ovarian puberty. These occur in the postnatal period: namely, the presence or absence of the ovaries and chronic treatment with estradiol. Because the females in the neuroendocrine model had their ovaries removed during the first few weeks of life, it is theoretically possible that the ovary secretes a factor(s) that obviates the fetal effects of androgen on the hypothalamus of the ovary-intact lamb, or this factor(s) may protect the brain from further damage that might occur in the postnatal period. However, what these substances might be and the nature of the underlying mechanism of action cannot even be predicted at present. A more probable explanation relates to the second difference, which is the pattern of estrogen experienced after birth, either the unvarying relatively high concentrations in the ovariectomized, estradiol-implanted animals or the low, fluctuating levels in the ovary-intact animals. Our unexpected results would be explained if after prenatal testosterone exposure the constant estrogen environment induced further disruption of the neural network controlling the GnRH surge, rendering the surge-generating mechanism nonfunctional in the neuroendocrine model. Indeed, we have some circumstantial support for this conjecture. Specifically, prepubertal, ovary-intact Suffolk females lambs that were exposed to chronic circulating concentrations of 13 pg/ml estradiol from 20 wk of age (SILASTIC capsule of estradiol) either failed to produce any estrous cycles (67%) or had a delayed onset of cycles in the first breeding season (11). Recent studies in rodents have shown that exposure of young cycling rats to chronic elevations in circulating estradiol abolishes steroid-induced LH surges over time (28), in part by inactivating GnRH neurons (29). Similar observations have been reported in cows (30), suggesting that the GnRH surge system is vulnerable to ovarian steroids in postnatal life. In view of the most interesting differences evident in the two models, it is clear that there is yet much to learn about how the prenatal and postnatal steroid environments interact to program development and function of the preovulatory GnRH surge system.
A second constant feature of numerous past studies on the timing of puberty in normal female lambs at Michigan and Babraham is that the escape from negative feedback in the ovariectomized, estrogen-implanted female at puberty is coincident with the onset of cycles in reproductive function in the ovary-intact animal. Thus, it was extremely surprising for us to find that neuroendocrine puberty preceded ovarian puberty by 15 or 21 wk, respectively, in lambs either androgenized for only the latter half of the critical period or for the entire critical period. Interestingly, this temporal separation of the times of neuroendocrine and ovarian puberty has also been observed within the same individual, as intact female lambs treated postnatally with chronic estrogen from 20 wk of age exhibit a pubertal LH rise long before ovulations begin (11). This further strengthens the possibility that postnatal exposure to unwavering estrogen concentrations disrupts the GnRH surge mechanism.
Another striking finding of this study is that fetal testosterone exposure severely disrupts the generation of reproductive cycles in the second breeding season. Specifically, none of the T60 ewes produced cycles, and fewer of the T30 ewes had normal, consecutive cycles than in the first breeding season. Furthermore, in the latter group the length of the second breeding season and the total number of cycles produced were significantly reduced compared with that in controls. This was in part due to the irregular nature of the progesterone cycles. These data indicate an inverse relationship between the length and/or level of exposure of the fetus to testosterone and the length of time before disruptive cycles become apparent. They also suggest that reproductive cycles in androgenized ewes become progressively disrupted with age. Interestingly, similar results have been reported in rats that were treated with low concentrations of testosterone propionate during the late postnatal critical period (23), a phenomenon that has been termed delayed anovulatory syndrome. Other data support our view that delayed anovulatory syndrome can be produced in androgenized ewes. In an early study by Clarke and colleagues (27), female lambs born to mothers treated with testosterone implants from d 50100 of pregnancy displayed fewer ovulations during the second breeding season (as assessed by laparotomy) than during the first (as assessed from plasma concentrations of progesterone collected over a period of 2535 d).
The reason for the progressive loss of fertility in the testosterone-treated female sheep is unknown, but it could originate from developmental changes occurring at one or more levels of the hypothalamic-pituitary-gonadal axis. Two main hypotheses may be put forward to explain the gradual decline in fertility. First, the GnRH neuronal network within the hypothalamus may become unresponsive to the positive feedback actions of estradiol with time. This situation may have been programmed by the androgen exposure in utero or exacerbated by the elevated circulating concentrations of LH found in these animals during postnatal life (9, 10, 31). A second testable hypothesis is that exposure to androgen in utero disrupts the functioning of the ovary such that ovarian development and/or folliculogenesis are abnormal in the androgenized ewe. Such effects may include prenatal androgen exposure altering the number of follicles present in the ovaries of androgenized lambs such that there are fewer follicles present in the ovary at birth. Another possibility is that follicle development is abnormal in the androgenized ewe, leading to compromised ovarian estradiol output. Relative to this, evidence from prepubertal androgenized lambs suggests that folliculogenesis is abnormal in these lambs, and this is manifest as enlarged, multifollicular ovaries (32, 33, 34, 35). It will be important in the future to determine the contributory roles of compromised estradiol feedback and ovarian disruption in facilitating the progressive loss of cyclicity, and studies that explore this are currently underway in our laboratories. Of interest is the observation that the compromised feedback and multifollicular morphology of ovaries from T60 ewes are remarkably similar to those found in women with disorders of androgen excess, such as polycystic ovary syndrome and congenital adrenal hyperplasia (36, 37, 38). Polycystic ovary syndrome is a disorder associated with abnormal follicle development, hyperandrogenization, and hypersecretion of LH and is probably the most common cause of anovulation in women of reproductive age (36). Many of these characteristics are also displayed by female sheep androgenized in utero (10, 32, 33, 34, 35, 39), raising the possibility that prenatal androgen exposure could be a developmental factor implicated in the etiology of this common disorder.
 |
Acknowledgments
|
|---|
We are extremely grateful to Andrew Dady, Tony Jones, Trevor Richter, and Martin White for assistance with the collection and processing of blood samples, surgery, and the administration of steroids, and to James Taylor for help with carrying out the progesterone assays. We thank the NIDDK for supplying both the LH for iodination and the LH standard.
 |
Footnotes
|
|---|
1 Present address: Department of Reproductive Science and Medicine, Institute of Reproductive and Developmental Biology, Imperial College London, Hammersmith Hospital, Du Cane Road, London W12 0NN, United Kingdom. 
This work was supported by the Biotechnology and Biological Sciences Research Council and Wellbeing.
Abbreviations: OVX+E, Ovariectomized and estrogen treated; T60, exposed to testosterone from d 3090; T90, exposed to testosterone from d 6090; TP, testosterone propionate.
Received October 25, 2002.
Accepted for publication January 6, 2003.
 |
References
|
|---|
- Short RV 1974 Sexual differentiation of the brain of sheep. In: Forest GM, Bertrand J, eds. The sexual endocrinology of the perinatal period. Lyon: Colloque International INSERM; 121142
- Clarke IJ, Scaramuzzi RJ, Short RV 1976 Effects of testosterone implants in pregnant ewes on their female offspring. J Embryol Exp Morphol 36:8799[Medline]
- Foster DL 1994 Puberty in the sheep. In: Knobil E and Neill JD, eds. The physiology of reproduction, vol 2. New York: Raven Press; 411451
- Wood RI, Foster DL 1998 Sexual differentiation of reproductive neuroendocrine function in sheep. Rev Reprod 3:130140[Abstract]
- Wood RI, Robinson JE, Forsdike RA, Padmanabhan V, Foster DL, Sexual differentiation of the timing of sexual maturation in sheep. 5th International Conference on the Control of the Onset of Puberty, Liege, Belgium, 1999, pp 269287
- Foster DL, Padmanabhan V, Wood RI, Robinson JE 2002 Sexual differentiation of the neuroendocrine control of gonadotropin secretion: concepts derived from the sheep. Reproduction 59(Suppl):8399
- Robinson JE, Birch RA, Taylor JA, Foster DL, Padmanabhan V 2002 In utero programming of sexually differentiated gonadotrophin releasing hormone (GnRH) secretion. Dom Anim Endocrinol 23:4353[CrossRef][Medline]
- Herbosa CG, Dahl GE, Evans NP, Pelt J, Wood RI, Foster DL 1996 Sexual differentiation of the surge mode of gonadotropin secretion: prenatal androgens abolish the gonadotropin-releasing hormone surge in the sheep. J Neuroendocrinol 8:627633[CrossRef][Medline]
- Sharma TP, Herkhimer C, West C, Ye W, Birch R, Robinson JE, Foster DL, Padmanabhan V 2002 Fetal programming: prenatal androgen disrupts positive feedback actions of estradiol but does not effect timing of puberty in female sheep. Biol Reprod 66:924933[Abstract/Free Full Text]
- Robinson JE, Taylor JA, Forsdike RA 1999 In utero exposure of female lambs to testosterone reduces the sensitivity of the GnRH neuronal network to inhibition by progesterone. Endocrinology 140:57975805[Abstract/Free Full Text]
- Foster DL, Karsch FJ, Olster DH, Ryan KD, Yellon SM 1986 Determinants of puberty in a seasonal breeder. Recent Prog Horm Res 42:331384[Medline]
- Ebling FJP, Foster DL 1988 Photoperiod requirements for puberty differ from those for the onset of the adult breeding season in the female sheep. J Reprod Fertil 84:283293[Abstract/Free Full Text]
- Wood RI, Ebling FJP, IAnson H, Foster DL 1991 Prenatal androgens time neuroendocrine sexual maturation. Endocrinology 128:24572468[Abstract]
- Karsch FJ, Legan SJ, Ryan KD, Foster DL 1980 Importance of estradiol and progesterone in regulating LH secretion and estrous behaviour during the sheep estrous cycle. Biol Reprod 23:404413[Abstract]
- Harris TG, Robinson JE, Evans NE, Skinner DC, Herbison AE 1998 Gonadotropin-releasing hormone messenger ribonucleic acid expression changes before the onset of the estradiol-induced luteinizng hormone surge in the ewe. Endocrinology 139:5764[Abstract/Free Full Text]
- Masek KS, Wood RI, Foster DL 1999 Prenatal dihydrotestosterone differentially masculinizes tonic and surge modes of luteinizing hormone secretion in sheep. Endocrinology 140:34593466[Abstract/Free Full Text]
- Niswender GD, Reichert Jr LE, Midgley Jr AR, Nalbandov AV 1969 Radioimmunoassay for bovine and ovine luteinizing hormone. Endocrinology 84:11661173[Medline]
- Robinson JE, Kendrick KM 1992 Inhibition of luteinizing hormone secretion in the ewe by progesterone: associated changes in the release of
-aminobutyric acid and noradrenaline in the preoptic area as measured by intracranial microdialysis. J Neuroendocrinol 4:231236[CrossRef]
- Padmanabhan V, Evans NP, Dahl GE, McFadden KL, Mauger DT, Karsch FJ 1995 Evidence for short or ultrashort loop negative feedback of gonadotropin-releasing hormone secretion. Neuroendocrinology 62:248258[Medline]
- Goy RW, McEwen BS 1980 Sexual differentiation of the brain. Cambridge: MIT Press
- Resko JA, Roselli CE 1997 Prenatal hormones organise sex differences in the neuroendocrine reproductive system; observations on guinea pigs and non human primates. Cell Mol Neurobiol 17:627648[CrossRef][Medline]
- Herman RA, Jones B, Mann DR, Wallen K 2000 Timing of prenatal androgen exposure: anatomical and endocrine effects on juvenile male and female rhesus monkeys. Horm Behav 38:5266[CrossRef][Medline]
- Gorski RA 1968 Influence of age on the response to paranatal administration of a low dose of androgen. Endocrinology 82:10011004[Medline]
- Abbott DH, Dumesic DA, Eisner JR, Kemnitz JW, Goy RW 1997 The prenatally androgenised female rhesus monkey as a model for PCOS. In: Azziz R, Nestler JE, Dweailly D, eds. Androgen excess disorders in women. Philadelphia: Lippincott-Raven; 369382
- Herbosa CG, Wood RI, Foster DL 1995 Prenatal androgens modify the reproductive response to photoperiod in the developing sheep. Biol Reprod 52:163169[Abstract]
- Herbosa CG, Foster DL 1996 Defeminization of the reproductive response to photoperiod occurs early in prenatal development in the sheep. Biol Reprod 54:420428[Abstract]
- Clarke IJ, Scaramuzzi RJ, Short RV 1977 Ovulation in prenatally androgenized ewes. J Endocrinol 73:385389[Abstract/Free Full Text]
- Tsai HW, Legan SJ 2001 Chronic elevation of estradiol in young ovariectomized rats causes aging-like loss of steroid-induced luteinizing hormone surges. Biol Reprod 64:684688[Abstract/Free Full Text]
- Tsai HW, Legan SJ 2002 Loss of luteinizing hormone surges induced by chronic estradiol is associated with decreased activation of gonadotropin-releasing hormone neurons. Biol Reprod 66:11041110[Abstract/Free Full Text]
- Nanda AS, Ward WR, Dobson H 1991 Lack of LH response to estradiol treatment in cows with cystic ovarian disease and effect of progesterone treatment or manual rupture. Res Vet Sci 51:180184[Medline]
- Fabre-Nys C, Venier G 1991 Sexual differentiation of sexual behaviour and preovulatory LH surge in ewes. Psychoneuroendocrinology 16:383396[CrossRef][Medline]
- Padmanabham V, Evans NP, Taylor JA, Robinson JE 1997 Prenatal exposure to androgens leads to the development of cystic ovaries in the sheep. Biol Reprod 56:194[Abstract]
- Forsdike RA, Taylor JA, Richter TA, Robinson JE 1999 Is the prenatally androgenized sheep a good model for polycystic ovary syndrome (PCOS) in women [Abstract 23]? J Reprod Fertil Abstr Ser 25:24
- West CR, Foster D, Padmanabhan V 2001 Intra-follicular activin availability is altered in prenatally-androgenized lambs. Mol Cell Endocrinol 185:5159[CrossRef][Medline]
- Birch RA, Robinson JE, Hardy K, Franks S 2002 Prenatal follicle distribution is abnormal in prenatally androgenized ewes [Abstract 41]. Reprod Abstr Ser 28:19
- Franks S 1995 Polycystic ovary syndrome. N Engl J Med 333:853861[Free Full Text]
- Hague WM, Adams J, Rodda C, Brook CG, de Bruyn R, Grant DB, Jacobs HS 1994 The prevalence of polycystic ovaries in patients with congenital adrenal hyperplasia and their close relatives. Clin Endocrinol (Oxf) 33:501510
- Barnes RB, Rosenfield RL, Erhmann DA, Cara JF, Cuttler L, Levitsky LL, Rosenthal IM 1994 Ovarian hyperandrogenism as a result of congenital adrenal virilizing disorders: evidence for perinatal masculinization of neuroendocrine function in women. J Clin Endocrinol Metab 79:13281333[Abstract]
- Abbott DH, Dumesic DA, Franks S 2002 Developmental origins of polycystic ovary syndrome: a hypothesis. J Endocrinol 174:15[Abstract]
This article has been cited by other articles:

|
 |

|
 |
 
T. L. Steckler, J. S. Lee, W. Ye, E. K. Inskeep, and V. Padmanabhan
Developmental Programming: Exogenous Gonadotropin Treatment Rescues Ovulatory Function But Does Not Completely Normalize Ovarian Function in Sheep Treated Prenatally with Testosterone
Biol Reprod,
October 1, 2008;
79(4):
686 - 695.
[Abstract]
[Full Text]
[PDF]
|
 |
|

|
 |

|
 |
 
L. M. Jackson, K. M. Timmer, and D. L. Foster
Sexual Differentiation of the External Genitalia and the Timing of Puberty in the Presence of an Antiandrogen in Sheep
Endocrinology,
August 1, 2008;
149(8):
4200 - 4208.
[Abstract]
[Full Text]
[PDF]
|
 |
|

|
 |

|
 |
 
L. M. Garcia-Segura, B. Lorenz, and L. L DonCarlos
The role of glia in the hypothalamus: implications for gonadal steroid feedback and reproductive neuroendocrine output
Reproduction,
April 1, 2008;
135(4):
419 - 429.
[Abstract]
[Full Text]
[PDF]
|
 |
|

|
 |

|
 |
 
M. Manikkam, R. C. Thompson, C. Herkimer, K. B. Welch, J. Flak, F. J. Karsch, and V. Padmanabhan
Developmental Programming: Impact of Prenatal Testosterone Excess on Pre- and Postnatal Gonadotropin Regulation in Sheep
Biol Reprod,
April 1, 2008;
78(4):
648 - 660.
[Abstract]
[Full Text]
[PDF]
|
 |
|

|
 |

|
 |
 
A. Veiga-Lopez, W. Ye, D.J. Phillips, C. Herkimer, P.G. Knight, and V. Padmanabhan
Developmental Programming: Deficits in Reproductive Hormone Dynamics and Ovulatory Outcomes in Prenatal, Testosterone-Treated Sheep
Biol Reprod,
April 1, 2008;
78(4):
636 - 647.
[Abstract]
[Full Text]
[PDF]
|
 |
|

|
 |

|
 |
 
C. E. Roselli, H. Stadelman, R. Reeve, C. V. Bishop, and F. Stormshak
The Ovine Sexually Dimorphic Nucleus of the Medial Preoptic Area Is Organized Prenatally by Testosterone
Endocrinology,
September 1, 2007;
148(9):
4450 - 4457.
[Abstract]
[Full Text]
[PDF]
|
 |
|

|
 |

|
 |
 
T. Steckler, M. Manikkam, E. K. Inskeep, and V. Padmanabhan
Developmental Programming: Follicular Persistence in Prenatal Testosterone-Treated Sheep Is Not Programmed by Androgenic Actions of Testosterone
Endocrinology,
July 1, 2007;
148(7):
3532 - 3540.
[Abstract]
[Full Text]
[PDF]
|
 |
|

|
 |

|
 |
 
R. A Forsdike, K. Hardy, L. Bull, J. Stark, L. J Webber, S. Stubbs, J. E Robinson, and S. Franks
Disordered follicle development in ovaries of prenatally androgenized ewes
J. Endocrinol.,
February 1, 2007;
192(2):
421 - 428.
[Abstract]
[Full Text]
[PDF]
|
 |
|

|
 |

|
 |
 
M. Savabieasfahani, K. Kannan, O. Astapova, N. P. Evans, and V. Padmanabhan
Developmental Programming: Differential Effects of Prenatal Exposure to Bisphenol-A or Methoxychlor on Reproductive Function
Endocrinology,
December 1, 2006;
147(12):
5956 - 5966.
[Abstract]
[Full Text]
[PDF]
|
 |
|

|
 |

|
 |
 
K. D. Malcolm, L. M. Jackson, C. Bergeon, T. M. Lee, V. Padmanabhan, and D. L. Foster
Long-Term Exposure of Female Sheep to Physiologic Concentrations of Estradiol: Effects on the Onset and Maintenance of Reproductive Function, Pregnancy, and Social Development in Female Offspring
Biol Reprod,
December 1, 2006;
75(6):
844 - 852.
[Abstract]
[Full Text]
[PDF]
|
 |
|

|
 |

|
 |
 
J. Robinson
Prenatal programming of the female reproductive neuroendocrine system by androgens.
Reproduction,
October 1, 2006;
132(4):
539 - 547.
[Abstract]
[Full Text]
[PDF]
|
 |
|

|
 |

|
 |
 
S M Carlsen, G Jacobsen, and P Romundstad
Maternal testosterone levels during pregnancy are associated with offspring size at birth.
Eur. J. Endocrinol.,
August 1, 2006;
155(2):
365 - 370.
[Abstract]
[Full Text]
[PDF]
|
 |
|

|
 |

|
 |
 
M. Manikkam, T. L. Steckler, K. B. Welch, E. K. Inskeep, and V. Padmanabhan
Fetal Programming: Prenatal Testosterone Treatment Leads to Follicular Persistence/Luteal Defects; Partial Restoration of Ovarian Function by Cyclic Progesterone Treatment
Endocrinology,
April 1, 2006;
147(4):
1997 - 2007.
[Abstract]
[Full Text]
[PDF]
|
 |
|

|
 |

|
 |
 
J. Pielecka, S. D. Quaynor, and S. M. Moenter
Androgens Increase Gonadotropin-Releasing Hormone Neuron Firing Activity in Females and Interfere with Progesterone Negative Feedback
Endocrinology,
March 1, 2006;
147(3):
1474 - 1479.
[Abstract]
[Full Text]
[PDF]
|
 |
|

|
 |

|
 |
 
S. E. Recabarren, V. Padmanabhan, E. Codner, A. Lobos, C. Duran, M. Vidal, D. L. Foster, and T. Sir-Petermann
Postnatal developmental consequences of altered insulin sensitivity in female sheep treated prenatally with testosterone
Am J Physiol Endocrinol Metab,
November 1, 2005;
289(5):
E801 - E806.
[Abstract]
[Full Text]
[PDF]
|
 |
|

|
 |

|
 |
 
D.H. Abbott, D.K. Barnett, C.M. Bruns, and D.A. Dumesic
Androgen excess fetal programming of female reproduction: a developmental aetiology for polycystic ovary syndrome?
Hum. Reprod. Update,
July 1, 2005;
11(4):
357 - 374.
[Abstract]
[Full Text]
[PDF]
|
 |
|

|
 |

|
 |
 
T. Steckler, J. Wang, F. F. Bartol, S. K. Roy, and V. Padmanabhan
Fetal Programming: Prenatal Testosterone Treatment Causes Intrauterine Growth Retardation, Reduces Ovarian Reserve and Increases Ovarian Follicular Recruitment
Endocrinology,
July 1, 2005;
146(7):
3185 - 3193.
[Abstract]
[Full Text]
[PDF]
|
 |
|